Aims To evaluate the diagnostic performance and non-inferiority of an abbreviated, non–intravenous contrast-enhanced magnetic resonance enterography (MRE) protocol compared with a standard gadolinium-enhanced protocol for the assessment of Crohn's disease activity, extent and complications. Materials and Methods This retrospective, single-centre study included 50 consecutive patients (mean age 42.7±16.9 years) with suspected or established Crohn’s disease who underwent standard MRE. Two radiologists, blinded to clinical data and reference reports, independently evaluated abbreviated protocol sequences (BTFE, T2-weighted SPAIR and TSE, DWI/ADC). Diagnostic performance was assessed against the report as reference standard. Interobserver agreement for categorical variables was analysed using Cohen’s and Fleiss’ kappa coefficients, while continuous variables were evaluated using intraclass correlation coefficients (ICC), with 95% confidence intervals. Bland–Altman analysis was performed. Results Agreement for disease activity was almost perfect (κ = 0.83, 95% CI: 0.64–0.96)(Reader A: sensitivity 93.5% (95% CI: 78.6–99.2), specificity 84.2% (60.4–96.6); reader B: sensitivity 90.3% (74.2–98.0), specificity 94.7% (74.0–99.9)). Agreement for anatomical location was also almost perfect. Substantial agreement was observed for stenosis detection (κ from 0.66 (95% CI: 0.29–0.89) to 0.73 (0.44–0.92)), whereas agreement for ulceration and penetrating complications was lower and heterogeneous. Quantitative measurements demonstrated excellent reproducibility (ICC>0.87, 95% CI: 0.79–0.93). Conclusion The abbreviated, non–intravenous contrast-enhanced MRE protocol demonstrates non-inferior diagnostic performance for the evaluation of disease activity, extent and stenosis in Crohn’s disease. Its implementation may reduce examination time and contrast exposure, although contrast-enhanced imaging remains indicated for ulcerations and penetrating complications assessment.
原文链接
分享
引用
